The prototype wild type AAV genome (AAV type 2) is a 4,600 base single stranded DNA molecule including a terminal 145 base repeat sequence (ITR) at each end (Muzyczka, Curr. Top. Microbiol. Immunol. 158:97-129, 1992). The AAV non-structural and structural open reading frames (ORFs) responsible for replication and encapsidation of the genome are derived from alternatively spliced transcripts originating from three distinct promoters. The replication (Rep) proteins are designated as Rep 78/68 and Rep 52/40, initiated at the P5 and P19 promoters, respectively. The structural capsid (Cap) proteins (VP1, VP2, and VP3) are under the control of the P40 promoter (Muzyczka supra). A combination of the terminal repeats, cis-sequences proximal to the AAV promoters, Rep proteins, adenovirus gene products, and cellular factors are responsible for the appropriate expression of the AAV Rep and Cap proteins (Tratschin et al., Mol. Cell. Biol. 6:2884-2894 (1986); Beaton et al., J. Virol. 63:4450-4454 (1989); Chang et al., J. Virol. 63:3479-3488 (1989); McCarty et al., J. Virol. 65:2936-2945 (1991); Shi et al., Cell 67:377-388 (1991); Horer et al., J. Virol. 69:5484-5496 (1995); Kyostio et al., J. Virol. 69:6787-6796 (1995); Lewis et al., J. Virol. 69:1628-1636 (1995); Weger et al., J. Virol. 71:8437-8447 (1997); Ni et al., J. Virol. 72:2777-2787 (1998)).
Adeno-associated virus is a defective parvovirus that grows in cells in which certain functions are provided by a co-infecting helper virus. General reviews of AAV may be found in, for example, Carter, Handbook of Parvoviruses, Vol. I, pp. 169-228 (1989), and Berns, Virology, pp. 1743-1764, Raven Press, New York, N.Y. (1990), incorporated herein by reference. Examples of co-infecting viruses that provide helper functions for AAV growth and replication are adenoviruses, herpes viruses, and in some cases, poxviruses such as vaccinia. The nature of the helper function is not entirely understood, but it appears that the helper virus indirectly renders the cell permissive for AAV replication. This belief is supported by the observation that AAV replication may occur at low efficiency in the absence of helper virus co-infection if the cells are treated with agents that are either genotoxic or that disrupt the cell cycle.
Although AAV may replicate to a limited extent in the absence of helper virus in these unusual conditions, more generally infection of cells with AAV in the absence of helper functions results in the proviral AAV genome integrating into the host cell genome. If these cells are superinfected with a helper virus such as adenovirus, the integrated AAV genome can be rescued and replicated to yield a burst of infectious progeny AAV particles. The fact that integration of AAV appears to be efficient suggests that AAV would be a useful vector for introducing genes into cells for use, such as, in human gene therapy.
AAV has a very broad host range without any obvious species or tissue specificity and can replicate in virtually any cell line of human, simian or rodent origin provided that an appropriate helper is present. AAV is also relatively ubiquitous and has been isolated from a variety of animal species including most mammalian and several avian species.
AAV vectors including heterologous polynucleotide sequences flanked by the AAV terminal repeats (ITRs) can be assembled into virions when introduced into appropriate cells by transfection with packaging plasmids encoding the Rep and Cap reading frames, but lacking the ITRs, and by co-infection with the non-related helper adenovirus (Samulski et al., J. Virol. 63:3822-3828 (1989); Muzcyka, Curr. Top. Microbiol. Immunol. 158:97-129 (1992)).
Initial AAV packaging plasmids contained simple deletions of the packaging signal (ITRs) to supply the Rep and Cap proteins for AAV vector production without generating wild type virus (Samulski et al., supra (1989); Muzyczka, supra (1992)). The AAV/Ad packaging construct, for example, has been widely used to generate AAV vector stocks by co-transfection into human embryonic kidney 293 cells (Graham et al., J. Virol. 36:59-74 (1977); Samulski et al., J. Virol. 63:3822-3828 (1989)). A derivative packaging plasmid, ACG-2, was described which converted the Rep 79/68 initiation codon from AUG to ACG to reduce the level of Rep 78/68 expression (Li et al., J. Virol. 71:5236-5243 (1987)). Over expression of the Rep 78/68 proteins has been correlated with decreases in AAV vector production. Expression of the AAV reading frames from these plasmids is still dependent upon transcription from the AAV P5, P19, and P40 promoters.
Co-transfection of AAV/Ad and AAV vector plasmids has been shown to result in significant contamination of vector stocks with replication-competent AAV (rcAAV) presumably by recombination between the plasmids during transfection (Allen et al., J. Virol. 71:6816-6822 (1997); Salvetti et al., Hum. Gene Ther. 9:695-706 (1998); Wang et al., J. Virol. 72:5472-5480 (1998)). An AAV packaging plasmid has previously been described which split the AAV rep and cap genes, placing them in an inverted orientation relative to the wild type AAV, and replaced the P5 and P40 promoters with the mouse metallothionein and cytomegalovirus (CMV) regulatory sequences, respectively (MTrep/CMVcap, see FIG. 1 of Allen et al., J. Virol. 71:6816-6822 (1997); WO98/27204, each hereby incorporated herein by reference). AAV vector production with this packaging plasmid was comparable to levels of vector produced with the AAV/Ad packaging plasmid and was free of rcAAV. Production of AAV vector using MTrep/CMVcap was still enhanced by infection with helper adenovirus despite the substitution of the P5 and P40 promoter sequences.
Adenovirus infection is an easy and efficient method for supplying the helper activities necessary for AAV vector production. However, removal of adenovirus from AAV vector stocks is laborious and can result in the contamination of vector stocks with potentially immunoreactive adenovirus proteins. Recently it has been shown that the adenovirus activities required for AAV vector production can be supplied by transfection of 293 cells with plasmids containing the E2A, E4 and VA RNA genes (Xiao et al., Journal of Virology, 72:2224-2232 (1998); Matsushita et al., Gene Ther. 5:938-945 (1998)). Full helper activity required the transfection of all three transcription units (Matsushita et al., supra (1998), incorporated herein by reference).
There are at least two desirable features of any AAV vector designed for use in human gene therapy. First, the transducing vector must be generated at titers sufficiently high to be practicable as a delivery system. This is especially important for gene therapy strategies aimed at in vivo delivery of the vector. For example, it is likely that for many desirable applications of AAV vectors, such as treatment of cystic fibrosis by direct in vivo delivery to the airway, the required dose of transducing vector may be in excess of 1010 particles. Secondly, the vector preparations must be essentially free of wild type AAV virus (or any replication-competent AAV). The attainment of high titers of AAV vectors has been difficult for several reasons including preferential encapsidation of wild type AAV genomes (if they are present or generated by recombination), and the difficulty in generating sufficient complementing functions such as those provided by the wild type rep and cap genes. Useful cell lines expressing such complementing functions have been especially difficult to generate, in part because of pleiotropic inhibitory functions associated with the rep gene products. Thus, cells in which the rep gene is integrated and expressed tend to grow slowly or express Rep at very low levels.
Vectors based on adeno-associated viruses (AAV) have been used in many applications in vivo because they promote persistent gene expression in dividing and non-dividing cells in multiple somatic tissues of animals (see, e.g., Herzog R. W. et al., Nat Med 5:56-63 (1999); Passini M A et al., J Virol 77:7034-7040 (2003); Weber M et al., Mol Ther 7:774-781 (2003) and Mochizuki S et al., Gene Ther 11:1081-1086 (2004)). However, recent studies have shown that the lack of an immune response seen in many mouse and some large animal studies has not been duplicated in human trials. For example, in a clinical trial for hemophilia B, two of seven subjects given an AAV vector-expressing clotting factor IX (FIX) developed a transient self-limiting increase in liver transaminases, followed by clearance of the FIX-expressing cells at 4 to 8 weeks after delivery (Manno, C. S., et al., Nat. Med. 12:342-347 (2006)). In this clinical trial for hemophilia B, a cytotoxic lymphocyte (CTL) response to AAV capsid but not to FIX was detected in peripheral blood mononuclear cells. In another clinical trial involving lipoprotein lipase, clearance of AAV-transduced cells coincided with a CTL response towards the capsid, but not the transgene (Mingozzi, F., et al., Blood 114:2077-2086 (2009)).
Therefore, a need exists for generating AAV vector preparations that reduce or eliminate the immune response in a mammalian subject.